2123H vs 2122H in a 4343 monitor

[Guido]

Originally posted by Giskard
Did you try the notch filter?
Have you tried what Earl and Gordon suggested?
Try both of those suggestions and let us know?

I tried the notch and it just didn't sound right. I could not explain but the sound was bad.

I tried the 2,5mH Inductor and there was even more mids
Please somebody explain me that
Every crossover program prognose the mids to decrease and then they increase......

Finally I tried the 22 uF in parallel to the 20 uF.
A Wonder
The whole sound of the 4343 changed. It's much more soft now. I still A/B the one with additional cap to the one without and can't believe.

Thanks Earl for this. And please try to explain this to me.

The additional cap really is the ticket to implement the 2012H in the 4343.

[4313B]

Excellent!

Nice R & D Guido!

[Earl K]

Hi Guido

Glad it worked out. I still think the circuit needs more tweaking - but if you "don't think its broke then don't attempt to fix it".

Thanks Earl for this. And please try to explain this to me.

The idea is directly from JBLs own implemention of massaging/handling rising midband response on their deep-gapped speakers ( like your 2012, the ME150H, the 1500al, etc. ). Though the extra value in the cap "upsets" the classic LC ratios and their stability quotients - those rules seem to be able to be bent that much easier when dealing with transducers that are inherently better damped / or have that extra percentage of motional control from the motor topology .

I'm sure most software packages just can't account for this. The wizards at JBL designing these circuits know better than the standard text-books.

regards <. Earl K

[Earl K]

Hi Guido

I tried the 2,5mH Inductor and there was even more mids
Please somebody explain me that

The replacement of the original 1.7mh inductor with the 2.5mh inductor should have knocked down that 1K peak as GordonW suggested - if - they had similar DCR values.

So, what are the DC resistance values for these two chokes ?

I suspect your replacement was a couple of ohms lower.

If so, that will change a lot of things in this finicky circuit. For instance the preceeding high pass section in this MF band will be working with (seeing) the wrong impedance. This actually creates a new high-pass point around 200hz instead of the original 350hz ( based on the motional impedance figures I have from my mockup circuit ). With that 2012, you should hear that change if you are not biamping. I forget if you are.

Anyways, that's the best explanation I can offer without a 2122h and the original circuit in front of me.

regards <> Earl K

[Guido]

Originally posted by Earl K
Hi Guido
The replacement of the original 1.7mh inductor with the 2.5mh inductor should have knocked down that 1K peak as GordonW suggested - if - they had similar DCR values.
So, what are the DC resistance values for these two chokes ?

Mhmm

My inductor have 0,3 Ohm and the original have 0,6 Ohm. Should this be the reason?

[Earl K]

Hi Guido

The difference in the DCRs' would need to be greater than the .3 ohms you just quoted. Now, 2 ohms would be a significant change .

So differing DCRs for that part substitution isn't reason .

regards <> Earl K

[Ian Mackenzie]

I am planning a new diy version of the 3143 JBL for my brothers 4343 crossover using 8 ohm L Pads for the HF and UHF with fixed R1/R2 pads (not T section) . In order to get the most gain from bi amping I was planning to omit the midrange T section and its Pad.

Can anyone make any suggestions if the LC values will need adjustment?

Ian

[Earl K]

Hi Ian,

Can anyone make any suggestions if the LC values will need adjustment?

- Which LC values are you referring to ?

- Just the MF (midrange) leg of the crossover ?

- Or all of the legs ?

- I built up a portion of the (3143) MF last year and measured ( no documentation was done ) the AC impedances in various portions of the leg. The "T" pads do go a long way in "stabilizing" the working circuit impedances that the preceeding LC components will see ( even when those whacky 16 ohm Lpad are adjusted up or down ). These "TPads" help provide a meaningful foundation from which to calculate the necessary LC values - again, when using 16 ohm variable Lpads.

I think you ought to explore "why" 16 ohm Lpads were used in the first place with these transducers . Even a bit of conjecture in this area can be useful . As in; perhaps they provided better "damping" for the existing transducers .

- Moving to 8 ohm variable Lpads ( might ) need to be accompanied by a relocation of the conjugate inductors ( in the MF and HF sections) . You can look to the 4344mkII and 4348 tech sheets for inspiration in this area ( of inductor - relocation ). I use inductors in this position with my horn drivers - I feel they provide better "damping" in that position ( ie , after the resistors in the circuit ) .


regards <. Earl K

PS ; so I guess my answer is ; yes , some ( if not most ) LC values should be recalculated .

[4313B]

"I think you ought to explore "why" 16 ohm Lpads were used in the first place with these transducers ."

The 16 ohm L-Pads in the early 43xx monitors were used for power handling. Reference was usually at 12 o'clock so set the 16 ohm L-Pad there and measure the two legs. Use those values to rework your attenuation. To get it real close include the impedance loads of the drivers.

[Ian Mackenzie]

Thanks for that. I suppose I need to build the real networks and work it out from there.

I'll let you know what happens.


Ian

[ralphs99]

At least part of the reason of the rising response above 800Hz is due to the directivity index (DI) increase with increasing frequency. Most drivers are designed to produce flat power response which means a direct response that rises with frequency. The obvious exception are horn loaded transducers that feature constant directivity.

In order to model the driver in an analog simulator some account needs to taken of this effect, assuming the aim is to produce a flat direct amplitude response. But that depends upon the listening environment. If the room is very dead and the listening position is relatively close to the speakers, ie a near field type listening environment, the direct sound dominates in the midband and a flat direct amplitude response should be the target. If the room is quite live and the listening distance more than a few of meters from the speakers, the listening environment for the midband will be in the diffuse field, and the power response of the speakers comes more into play. I suspect that for most typical home environments something between the two will be most suitable. I have to stress that this effect is frequency dependant. The critical distance (ie the change from near field to far field) increases with decreasing frequency.

Guido: This may explain why the notch filters you have been trying are not sounding good to you. Notch filters are also problematic for the phase reponse. I suggest a simpler way around your problem may be to increase the toe-in of your speakers so that you are no longer directly on axis. 10-15 degrees should be enough. This solution works well for 4430's with constant directivity horns. For 43xx series monitors, the changing DI vs frequency of the exponential horns will negate the advantages of the extra toe-in to some degree ( HF loss). But it's worth a try as it's such a simple experiment.

Cheers, Ralph

[Ian Mackenzie]

I think I will build the 3145 network and be damned, I figure the only change will be the mid range LC values.


Ian

[Ian Mackenzie]

Heather,

Perhaps this thread may prove useful to you. There will be some degree of tweaking required . Regards all this humpty do, I measured the real gain of the mid filter passband quite sometime ago. The difference in spl of the mid cone driver was about 1 db between the biamp and full passive mode. That can be levelled with the L pads if you wish without stessing out over voltage drives too much. More important is balancing the driver outputs to blend and your taste. I find +- 1/3 db audible with four drivers.

There is so much play in those pads I wonder if they are a virtue or a sin. Regardless you will need to carefully balance all four drivers so they work as one!

Giskard is also discussing filters in another thead, an educational and insightful discussion for some and no doubt incomprehensible for others.


Ian

[4313B]

Giskard is also discussing filters in another thead, an educational and insightful discussion for some and no doubt incomprehensible for others.

http://www.audioheritage.org/vbullet...ad.php?t=13043

Ian - I deleted all my posts and will consider addressing your concerns in full in a new thread.

[hjames]

Thank you very much - I hadn't seen this thread before!
Even with the search engine there is so much that just doesn't show on this site! I often email myself bookmarks of useful threads when I find them - or when you heavy-hitters point them out to us! Again, much thanks!

I took a few glances at it and will definately read it in more detail over the weekend. Its Thanksgiving Day stateside and we've fixed a small feast to take to my dad (he's 84) about an hour away in Maryland. The family doesn't get together much like when we were younger, but if it makes him laugh and smile, I'm all for it.

I've plans to pull my stereo gear down off the media room shelving - I'll move the media shelves around and assemble the electronics in a rolling teak rack I used to use. Sad to say I did like the look of the beechwood shelf system (IKEA), but its just not conducive to the kind of periodic dinking I'm prone to do ... (not to be confused with the single-malt drinking I'm also prone to do).

Anyway - I've got to get the monitors down from "on high" and probably build some small platforms to get the slots at ear level. Once thats all arranged I'll take more pictures to post in another thread later.

Too much to do, but gotta run - have a great day! Thanks to everyone!

-= )-(eather =-

Heather,

Perhaps this thread may prove useful to you. There will be some degree of tweaking required . Regards all this humpty do, I measured the real gain of the mid filter passband quite sometime ago. The difference in spl of the mid cone driver was about 1 db between the biamp and full passive mode. That can be levelled with the L pads if you wish without stessing out over voltage drives too much. More important is balancing the driver outputs to blend and your taste. I find +- 1/3 db audible with four drivers.

There is so much play in those pads I wonder if they are a virtue or a sin. Regardless you will need to carefully balance all four drivers so they work as one!

Giskard is also discussing filters in another thead, an educational and insightful discussion for some and no doubt incomprehensible for others.


Ian